PRESCRIPTIVE ANALYTICS.

To better understand what prescriptive analytics is, we will use a relatable example.

Example

Imagine you are a barista working at a coffee shop, and you are responsible for purchasing coffee beans used at the coffee shop. On each Sunday morning over the past 2 years, you have been buying 7 kilograms of coffee beans that last you exactly for a week. During a festive week, as you report to work on a Wednesday morning. You realize that for the past three days, you have been using 2 kilograms of coffee beans daily and the trend will likely continue until the end of the week. From this, you can foresee that you are likely to finish your supply of coffee beans by noon. Now you are at a crossroads; you will either order more coffee beans or redirect the customers to a different coffee shop.

From the above coffee shop example, we notice three types of Business Analytics.

1.       Descriptive analytics – What happened in the past?

Ø  Using 7 kilograms of coffee beans weekly.

2.       Predictive analytics – What will happen in the future?

Ø  At noon there will probably be no coffee at the shop.

3.       Prescriptive analytics – What will you do?

Ø  Buy more coffee beans to meet the demand.

Ø  Redirect customers to another coffee shop

In 2021 Catherine Cote a marketing coordinator at Harvard Business School Online. Defined Prescriptive analytics as the process of using data to determine an optimal course of action. From the example, it is clear that if the goal is to make profits the optimal course of action would be to buy more coffee beans.

Optimization has grown and evolved from the traditional linear programming and metaheuristic methods taught in Statistics operation research classes into modern-day prescriptive analytics.

Tools and software for prescriptive analytics.

There are tools and software that can be used to automate and make the process of choosing the optimal solution easier and faster. Some of these tools include:

Ø  Optimization Software: IBM Decision Optimization, FICO Xpress.

Ø  Machine Learning Platforms: Python, R, DataRobot.

Ø  Data Visualization and Business Intelligence Tools: Tableau, Power BI.

Ø  Simulation Modeling Software: AnyLogic, Simio.

It is important to note that optimization done by software would still need human intelligence for judgment and final decision-making, ensuring the final decisions align with strategic goals and business values.

For such conversations and opportunities related to data science, analytics and Statistics let's start a conversation! Reach out to me via:

• Email: williammorara28@gmail.com
• LinkedIn: William Morara
• Phone: +254710489328